Upload
others
View
3
Download
0
Embed Size (px)
Citation preview
How to reduce (household) electricity use
1. Technological advancesMore efficient refrigerators, tv’s, etc.
2. Changing energy behaviourThis requires understanding energy behaviour
Special characteristics of psychological research
• Distinction between independent and dependent variables often not clear:everything affects everything
• Causal relations are hard. Often variables cause each other (A→ B and B→A)
• Many variables involved. Often many overlapping theories involved.• Technical background of psychologists more limited than of, e.g., biologists.Results should be interpretable for non-statisticians.
Solution: use psychological network models.
Psychological network models
In essence, psychological network models are:
A set of nodes and edges, depicting random variables and their relations.
Gaussian graphical models are a specific case of psychological network models.
Psychological networks vs. social networks
Networks
• Network of things
• Examples: social networks, trafficnetworks
• Nodes represent entities• Edges are either observed ordirectly derivable from observations
• Main goal: explainpresence/absence of edges,centrality, shortest paths
Psychological networks• Network of variables
• Examples: correlation networks,depression symptom models
• Nodes represent random variables• Edges are estimated statisticalrelations, subject to uncertainty
• Main goal: explain (co)occurrence ofcertain nodes
• Conceptually improvement to latentvariable models
(Table: Sacha Epskamp)
Psychological networks vs. social networks
Networks
• Network of things• Examples: social networks, trafficnetworks
• Nodes represent entities• Edges are either observed ordirectly derivable from observations
• Main goal: explainpresence/absence of edges,centrality, shortest paths
Psychological networks• Network of variables• Examples: correlation networks,depression symptom models
• Nodes represent random variables• Edges are estimated statisticalrelations, subject to uncertainty
• Main goal: explain (co)occurrence ofcertain nodes
• Conceptually improvement to latentvariable models
(Table: Sacha Epskamp)
Psychological networks vs. social networks
Networks
• Network of things• Examples: social networks, trafficnetworks
• Nodes represent entities
• Edges are either observed ordirectly derivable from observations
• Main goal: explainpresence/absence of edges,centrality, shortest paths
Psychological networks• Network of variables• Examples: correlation networks,depression symptom models
• Nodes represent random variables
• Edges are estimated statisticalrelations, subject to uncertainty
• Main goal: explain (co)occurrence ofcertain nodes
• Conceptually improvement to latentvariable models
(Table: Sacha Epskamp)
Psychological networks vs. social networks
Networks
• Network of things• Examples: social networks, trafficnetworks
• Nodes represent entities• Edges are either observed ordirectly derivable from observations
• Main goal: explainpresence/absence of edges,centrality, shortest paths
Psychological networks• Network of variables• Examples: correlation networks,depression symptom models
• Nodes represent random variables• Edges are estimated statisticalrelations, subject to uncertainty
• Main goal: explain (co)occurrence ofcertain nodes
• Conceptually improvement to latentvariable models
(Table: Sacha Epskamp)
Psychological networks vs. social networks
Networks
• Network of things• Examples: social networks, trafficnetworks
• Nodes represent entities• Edges are either observed ordirectly derivable from observations
• Main goal: explainpresence/absence of edges,centrality, shortest paths
Psychological networks• Network of variables• Examples: correlation networks,depression symptom models
• Nodes represent random variables• Edges are estimated statisticalrelations, subject to uncertainty
• Main goal: explain (co)occurrence ofcertain nodes
• Conceptually improvement to latentvariable models
(Table: Sacha Epskamp)
Psychological networks vs. social networks
Networks
• Network of things• Examples: social networks, trafficnetworks
• Nodes represent entities• Edges are either observed ordirectly derivable from observations
• Main goal: explainpresence/absence of edges,centrality, shortest paths
Psychological networks• Network of variables• Examples: correlation networks,depression symptom models
• Nodes represent random variables• Edges are estimated statisticalrelations, subject to uncertainty
• Main goal: explain (co)occurrence ofcertain nodes
• Conceptually improvement to latentvariable models
(Table: Sacha Epskamp)
Psychological networks
X ∼ N(µ,Σ), X : n× p
Example: grades for n = 44 students on p = 5 grades. Correlations:
mechanics vectors algebra analysis statisticsmechanics 1.00 0.32 0.41 0.38 0.27
vectors 0.32 1.00 0.43 0.23 0.22algebra 0.41 0.43 1.00 0.58 0.55analysis 0.38 0.23 0.58 1.00 0.51statistics 0.27 0.22 0.55 0.51 1.00
Psychological networks
X ∼ N(µ,Σ), X : n× pExample: grades for n = 88 students on p = 5 grades.Correlations (from Σ̂) and partial correlations (from Σ̂−1):
mch
vct
alganl
stt
mch
vct
alganl
stt
Psychological networks
X ∼ N(µ,Σ), X : n× pExample: grades for n = 88 students on p = 5 grades.Correlations (from Σ̂) and partial correlations (from Σ̂−1):
mch
vct
alganl
stt
mch
vct
alganl
stt
Psychological networks: fewer edges
I nodes→ I(I− 1)/2 potential edges. Quickly too many.
Graphical lasso. Estimate Σ−1 subject to L1-penalties. Commonly used is theGraphical lasso (Friedman, Hastie, Tibshirani, Biostatistics, 2008) approach withthe extended BIC to select tuning parameter.
mch
vct
alganl
stt
mch
vct
alganl
stt
Which nodes matter most?
mch
vct
alganl
stt
Betweenness Closeness Strength
−0.50 −0.25 0.00 0.25 0.50−1 0 1 −1 0 1
algebra
analysis
mechanics
statistics
vectors
Network Centrality Measures (standardised)
Wij: matrix with weights (∈ [−1, 1]) in the network. SDij: shortest distance fromnode i to node j.
• Node strength/centrality: Si =∑
jWij• Closeness: CLi = (
∑j SDij)−1
• Betweenness: Bi = #(paths through node i)/#(all paths)
Comparing networks
mch
vct
alganl
stt
mch
vct
alganl
stt
Below median score for mechanics Above median score for mechanics
• Structural Hamming Distance.Count number of edges that (dis)appeared/changed sign.Here: 5 out of 10.
• Network Comparison Test (Van Borculo et al.)Statistical permutation test, similar to Mantel’s test (1967).Here: p = .62.
Comparing networks
mch
vct
alganl
stt
mch
vct
alganl
stt
Below median score for mechanics Above median score for mechanics
• Structural Hamming Distance.Count number of edges that (dis)appeared/changed sign.Here: 5 out of 10.
• Network Comparison Test (Van Borculo et al.)Statistical permutation test, similar to Mantel’s test (1967).Here: p = .62.
Psychological networks: Beyond the basics
In this talk, only undirected GGM.
Network models also possible for:
• Directed graphs, when visualising temporal dynamics (e.g. VAR-models;Bringmann et al.) or causal models (DAGs).
• Ordinal data: polychoric rather than Pearson correlations.• Binary data: Ising-models.
Furthermore, all estimates (e.g. edge weights) can be equipped by bootstrap CI’s(Epskamp et al, 2017).
Buurkracht: Introduction
• Community energy initiative for promoting sustainable energy behaviour.• Research on energy behaviour usually focused on individual.• Community efforts might be more effective.• Combination of various psychological and sociological theories at play.
• N1 = 334 initiative participants, N2 = 360 right-door-neighbours.• Data on 22 variables, and 65 items, 7-point Likert scales.
Buurkracht: Introduction
• Community energy initiative for promoting sustainable energy behaviour.• Research on energy behaviour usually focused on individual.• Community efforts might be more effective.• Combination of various psychological and sociological theories at play.
• N1 = 334 initiative participants, N2 = 360 right-door-neighbours.• Data on 22 variables, and 65 items, 7-point Likert scales.
Buurkracht: Introduction
• Community energy initiative for promoting sustainable energy behaviour.• Research on energy behaviour usually focused on individual.• Community efforts might be more effective.• Combination of various psychological and sociological theories at play.
• N1 = 334 initiative participants, N2 = 360 right-door-neighbours.• Data on 22 variables, and 65 items, 7-point Likert scales.
Buurkracht: Variables in the study
Personal factors Values. Environmental self-identity. Personal importance ofsustainable energy behaviour. Outcome efficacy.
Social context Need to belong Need to be unique. Neighbourhood identification.Neighbourhood homogeneity. Neighbourhood interaction. Neighbourhoodenvironmental identity. Neighbourhood importance of sustainable energy behaviour.
Opinions on energy companies and the government Group-based anger.
Sustainable energy intentions and behaviour Household sustainable energyintentions. Collective sustainable energy intentions. Collective social intentions.Self-reported sustainable energy behaviour.
Initiative membership
Buurkracht: Findings
●●
●●
●● ●●
●●●●
●●●●●●
●●●●
●●
●●
●●●●●●●●
●●●●
●●
●● ●●●●
●●●● ●●
●●
●●●●
●●●●●●
●●●●
●●●●
●●
●●●● ●●
●●●●
●● ●●
●●
●●●●
●●●● ●●
●●●●●●
●●
●●●●
●●
●●
●●●●●●
●●
●●●●
●●
1
2
3
4
5
6
78
9
10
11
12
13
1415
16
17
18
19
20
2122
23
24
25 26
27
2829
3031
32
33
34
35
36
37
3839
40
41
42
43 44
45
46
47
48
49 50
515253
54
55
56
57
58
59
60
61
62
63
64
65
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
Altruistic valuesBiospheric ValuesCollective energy intentionsCollective social intentionsEgoistic ValuesEnvironmental Self_identityGroup−based angerHedonic ValuesHomogeneity in neighbourhoodIndividual energy intentionsInitiative outcome efficacyInteraction in neighbourhoodInteraction with neighbourhoodMembershipNeed to be UniqueNeed to belong Neighbourhood energy normsNeighbourhood environmental IdentityNeighbourhood identificationPast energy behaviorPersonal Energy NormsPersonal outcome efficacy
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
Altruistic valuesBiospheric ValuesCollective energy intentionsCollective social intentionsEgoistic ValuesEnvironmental Self_identityGroup−based angerHedonic ValuesHomogeneity in neighbourhoodIndividual energy intentionsInitiative outcome efficacyInteraction in neighbourhoodInteraction with neighbourhoodMembershipNeed to be UniqueNeed to belong Neighbourhood energy normsNeighbourhood environmental IdentityNeighbourhood identificationPast energy behaviorPersonal Energy NormsPersonal outcome efficacy
Buurkracht: Findings
12
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
Factors related to the social context5: Neighbourhood environmental Identity6: Neighbourhood identification7: Interaction with neighbourhood8: Neighbourhood energy norms9: Homogeneity in neighbourhood10: Interaction in neighbourhood13: Need to belong15: Need to be Unique20: Initiative outcome efficacy
Membership22: Membership
Opinions on energy companies21: Group based anger
Personal factors 11: Personal outcome efficacy12: Environmental Self identity14: Personal Energy Norms16: Altruistic values17: Biospheric Values18: Egoistic Values19: Hedonic Values
Sustainable energy intentions1: Collective energy intentions2: Collective social intentions3: Individual energy intentions4: Past energy behavior
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
Factors related to the social context5: Neighbourhood environmental Identity6: Neighbourhood identification7: Interaction with neighbourhood8: Neighbourhood energy norms9: Homogeneity in neighbourhood10: Interaction in neighbourhood13: Need to belong15: Need to be Unique20: Initiative outcome efficacy
Membership22: Membership
Opinions on energy companies21: Group based anger
Personal factors 11: Personal outcome efficacy12: Environmental Self identity14: Personal Energy Norms16: Altruistic values17: Biospheric Values18: Egoistic Values19: Hedonic Values
Sustainable energy intentions1: Collective energy intentions2: Collective social intentions3: Individual energy intentions4: Past energy behavior
Buurkracht: Findings
1
2
3
4
5
6
7
8
910
11
12
13
14
15
16
17
18
19
20
21
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
Factors related to the social context5: Neighbourhood environmental Identity6: Neighbourhood identification7: Interaction with neighbourhood8: Neighbourhood energy norms9: Homogeneity in neighbourhood10: Interaction in neighbourhood13: Need to belong15: Need to be Unique20: Initiative outcome efficacy
Opinions on energy companies21: Group based anger
Personal factors 11: Personal outcome efficacy12: Environmental Self identity14: Personal Energy Norms16: Altruistic values17: Biospheric Values18: Egoistic Values19: Hedonic Values
Sustainable energy intentions1: Collective energy intentions2: Collective social intentions3: Individual energy intentions4: Past energy behavior
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
Factors related to the social context5: Neighbourhood environmental Identity6: Neighbourhood identification7: Interaction with neighbourhood8: Neighbourhood energy norms9: Homogeneity in neighbourhood10: Interaction in neighbourhood13: Need to belong15: Need to be Unique20: Initiative outcome efficacy
Opinions on energy companies21: Group based anger
Personal factors 11: Personal outcome efficacy12: Environmental Self identity14: Personal Energy Norms16: Altruistic values17: Biospheric Values18: Egoistic Values19: Hedonic Values
Sustainable energy intentions1: Collective energy intentions2: Collective social intentions3: Individual energy intentions4: Past energy behavior
1
2
3
4
5
6
7
8
910
11
12
13
14
15
16
17
18
19
20
21
Structural Hamming Distance: 12
ESS: Introduction
• Open Data, www.europeansocialsurvey.org• “Public Attitudes to Climate Change, Energy Security, and Energy Preferences”• N = 38, 437 participants from 18 countries (880 to 2,766 per country).
Research questions:
1. What, and how strong, are the relations between the variables?2. Are these relations the same across countries?
ESS: Introduction
• Open Data, www.europeansocialsurvey.org• “Public Attitudes to Climate Change, Energy Security, and Energy Preferences”• N = 38, 437 participants from 18 countries (880 to 2,766 per country).
Research questions:
1. What, and how strong, are the relations between the variables?2. Are these relations the same across countries?
ESS: Introduction
• Open Data, www.europeansocialsurvey.org• “Public Attitudes to Climate Change, Energy Security, and Energy Preferences”• N = 38, 437 participants from 18 countries (880 to 2,766 per country).
Research questions:
1. What, and how strong, are the relations between the variables?2. Are these relations the same across countries?
ESS: the Module
The module “Public Attitudes to Climate Change, Energy Security, and EnergyPreferences” includes 32 items in the areas:
1. Beliefs on climate change2. Concerns about climate change and energy security3. Personal norms, efficacy and trust4. Energy preferences.
ESS: findings
EDM1
EDM2
EB1
ESP1
ESP2
ESP3ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
Policy supportPS1: Increasing taxes on fossil fuelsPS2: Subsidise renewable energyPS3: Ban least energy efficient appliances
Climate change beliefsCCB1: Belief in climate changeCCB2: Climate change causeCCB3: Impact of climate change
Climate change salienceCCS: Thought about climate change before today
Climate concernCC: Climate change worry
Efficacy beliefsEB1: Could use less energyEB2: Peoples impact on climate changeEB3: Likeliness of people limiting energy useEB4: Likeliness of governments limiting energy useEB5: My impact on climate change
Energy demand measuresEDM1: Buy energy efficient appliancesEDM2: Reduce your energy use
Energy security concernESC1: Power cutsESC2: ExpensiveESC3: Energy importsESC4: Fossil fuelsESC5: Natural disastersESC6: Insufficient powerESC7: Technical failuresESC8: Terrorist attacks
Energy supply source preferenceESP1: CoalESP2: Natural gasESP3: Hydroelectric powerESP4: Nuclear powerESP5: Sun or solar powerESP6: Wind powerESP7: Biomass energy
Personal normsPN: Personal responsibility to reduce climate change
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
Policy supportPS1: Increasing taxes on fossil fuelsPS2: Subsidise renewable energyPS3: Ban least energy efficient appliances
Climate change beliefsCCB1: Belief in climate changeCCB2: Climate change causeCCB3: Impact of climate change
Climate change salienceCCS: Thought about climate change before today
Climate concernCC: Climate change worry
Efficacy beliefsEB1: Could use less energyEB2: Peoples impact on climate changeEB3: Likeliness of people limiting energy useEB4: Likeliness of governments limiting energy useEB5: My impact on climate change
Energy demand measuresEDM1: Buy energy efficient appliancesEDM2: Reduce your energy use
Energy security concernESC1: Power cutsESC2: ExpensiveESC3: Energy importsESC4: Fossil fuelsESC5: Natural disastersESC6: Insufficient powerESC7: Technical failuresESC8: Terrorist attacks
Energy supply source preferenceESP1: CoalESP2: Natural gasESP3: Hydroelectric powerESP4: Nuclear powerESP5: Sun or solar powerESP6: Wind powerESP7: Biomass energy
Personal normsPN: Personal responsibility to reduce climate change
ESS: findings
EDM1
EDM2
EB1
ESP1
ESP2
ESP3
ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
EDM1
EDM2
EB1
ESP1
ESP2
ESP3
ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
EDM1
EDM2
EB1
ESP1
ESP2
ESP3
ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
EDM1
EDM2
EB1
ESP1
ESP2
ESP3
ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
EDM1
EDM2
EB1
ESP1
ESP2
ESP3
ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
EDM1
EDM2
EB1
ESP1
ESP2
ESP3
ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
EDM1
EDM2
EB1
ESP1
ESP2
ESP3
ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
EDM1
EDM2
EB1
ESP1
ESP2
ESP3
ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
EDM1
EDM2
EB1
ESP1
ESP2
ESP3
ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
EDM1
EDM2
EB1
ESP1
ESP2
ESP3
ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
EDM1
EDM2
EB1
ESP1
ESP2
ESP3
ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
EDM1
EDM2
EB1
ESP1
ESP2
ESP3
ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
EDM1
EDM2
EB1
ESP1
ESP2
ESP3
ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
EDM1
EDM2
EB1
ESP1
ESP2
ESP3
ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
EDM1
EDM2
EB1
ESP1
ESP2
ESP3
ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
EDM1
EDM2
EB1
ESP1
ESP2
ESP3
ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
EDM1
EDM2
EB1
ESP1
ESP2
ESP3
ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
EDM1
EDM2
EB1
ESP1
ESP2
ESP3
ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
EDM1
EDM2
EB1
ESP1
ESP2
ESP3
ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
EDM1
EDM2
EB1
ESP1
ESP2
ESP3
ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
EDM1
EDM2
EB1
ESP1
ESP2
ESP3
ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
EDM1
EDM2
EB1
ESP1
ESP2
ESP3
ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
No major between-network differences. K-means clustering yields 1 cluster.
ESS: findings
−2
0
2
EDM
1ED
M2
EB1
ESP1
ESP2
ESP3
ESP4
ESP5
ESP6
ESP7
ESC
1ES
C2
ESC
3ES
C4
ESC
5ES
C6
ESC
7ES
C8
CC
B1C
CS
CC
B2 PN CC
CC
B3 EB2
EB3
EB4
EB5
PS1
PS2
PS3
Cen
tral
ityCountry
Austra
Belgium
Czech Republic
Estonia
Finland
France
Germany
Iceland
Ireland
Israel
Netherlands
Norway
Poland
Russia
Slovenia
Sweden
Switzerland
United Kingdom
ESS: zooming in
EDM1
EDM2
EB1
ESP1
ESP2
ESP3ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
Policy supportPS1: Increasing taxes on fossil fuelsPS2: Subsidise renewable energyPS3: Ban least energy efficient appliances
Climate change beliefsCCB1: Belief in climate changeCCB2: Climate change causeCCB3: Impact of climate change
Climate change salienceCCS: Thought about climate change before today
Climate concernCC: Climate change worry
Efficacy beliefsEB1: Could use less energyEB2: Peoples impact on climate changeEB3: Likeliness of people limiting energy useEB4: Likeliness of governments limiting energy useEB5: My impact on climate change
Energy demand measuresEDM1: Buy energy efficient appliancesEDM2: Reduce your energy use
Energy security concernESC1: Power cutsESC2: ExpensiveESC3: Energy importsESC4: Fossil fuelsESC5: Natural disastersESC6: Insufficient powerESC7: Technical failuresESC8: Terrorist attacks
Energy supply source preferenceESP1: CoalESP2: Natural gasESP3: Hydroelectric powerESP4: Nuclear powerESP5: Sun or solar powerESP6: Wind powerESP7: Biomass energy
Personal normsPN: Personal responsibility to reduce climate change
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
Policy supportPS1: Increasing taxes on fossil fuelsPS2: Subsidise renewable energyPS3: Ban least energy efficient appliances
Climate change beliefsCCB1: Belief in climate changeCCB2: Climate change causeCCB3: Impact of climate change
Climate change salienceCCS: Thought about climate change before today
Climate concernCC: Climate change worry
Efficacy beliefsEB1: Could use less energyEB2: Peoples impact on climate changeEB3: Likeliness of people limiting energy useEB4: Likeliness of governments limiting energy useEB5: My impact on climate change
Energy demand measuresEDM1: Buy energy efficient appliancesEDM2: Reduce your energy use
Energy security concernESC1: Power cutsESC2: ExpensiveESC3: Energy importsESC4: Fossil fuelsESC5: Natural disastersESC6: Insufficient powerESC7: Technical failuresESC8: Terrorist attacks
Energy supply source preferenceESP1: CoalESP2: Natural gasESP3: Hydroelectric powerESP4: Nuclear powerESP5: Sun or solar powerESP6: Wind powerESP7: Biomass energy
Personal normsPN: Personal responsibility to reduce climate change
Detailed look at five variables
Werfel, 2017; Noblet & McCoy, 2017:
Engaging inEnergy Behaviour
Support forEnergy Policies
Steinhorst & Matthies, 2016; Thøgersen & Noblet, 2012:
Engaging inEnergy Behaviour
Support forEnergy Policies
Desire for consistency:Stronger relation between related concepts
Detailed look at five variables
B1B2
P1
P2
P3
●
●
●
●
●
BehaviourB1: Buying an energy−efficient applianceB2: Energy−saving behaviour
Policy supportP1: Taxes on fossil fuelsP2: Subsidies for renewable energyP3: Ban on inefficient appliances
●
●
●
●
●
BehaviourB1: Buying an energy−efficient applianceB2: Energy−saving behaviour
Policy supportP1: Taxes on fossil fuelsP2: Subsidies for renewable energyP3: Ban on inefficient appliances
B1B2
P1
P2
P3
●
●
●
●
●
BehaviourB1: Buying an energy−efficient applianceB2: Energy−saving behaviour
Policy supportP1: Taxes on fossil fuelsP2: Subsidies for renewable energyP3: Ban on inefficient appliances
●
●
●
●
●
BehaviourB1: Buying an energy−efficient applianceB2: Energy−saving behaviour
Policy supportP1: Taxes on fossil fuelsP2: Subsidies for renewable energyP3: Ban on inefficient appliances
ESS: zooming in
EDM1
EDM2
EB1
ESP1
ESP2
ESP3ESP4
ESP5
ESP6
ESP7
ESC1
ESC2
ESC3
ESC4
ESC5
ESC6
ESC7
ESC8
CCB1
CCS
CCB2
PN
CC
CCB3
EB2
EB3
EB4
EB5
PS1
PS2
PS3
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
Policy supportPS1: Increasing taxes on fossil fuelsPS2: Subsidise renewable energyPS3: Ban least energy efficient appliances
Climate change beliefsCCB1: Belief in climate changeCCB2: Climate change causeCCB3: Impact of climate change
Climate change salienceCCS: Thought about climate change before today
Climate concernCC: Climate change worry
Efficacy beliefsEB1: Could use less energyEB2: Peoples impact on climate changeEB3: Likeliness of people limiting energy useEB4: Likeliness of governments limiting energy useEB5: My impact on climate change
Energy demand measuresEDM1: Buy energy efficient appliancesEDM2: Reduce your energy use
Energy security concernESC1: Power cutsESC2: ExpensiveESC3: Energy importsESC4: Fossil fuelsESC5: Natural disastersESC6: Insufficient powerESC7: Technical failuresESC8: Terrorist attacks
Energy supply source preferenceESP1: CoalESP2: Natural gasESP3: Hydroelectric powerESP4: Nuclear powerESP5: Sun or solar powerESP6: Wind powerESP7: Biomass energy
Personal normsPN: Personal responsibility to reduce climate change
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
Policy supportPS1: Increasing taxes on fossil fuelsPS2: Subsidise renewable energyPS3: Ban least energy efficient appliances
Climate change beliefsCCB1: Belief in climate changeCCB2: Climate change causeCCB3: Impact of climate change
Climate change salienceCCS: Thought about climate change before today
Climate concernCC: Climate change worry
Efficacy beliefsEB1: Could use less energyEB2: Peoples impact on climate changeEB3: Likeliness of people limiting energy useEB4: Likeliness of governments limiting energy useEB5: My impact on climate change
Energy demand measuresEDM1: Buy energy efficient appliancesEDM2: Reduce your energy use
Energy security concernESC1: Power cutsESC2: ExpensiveESC3: Energy importsESC4: Fossil fuelsESC5: Natural disastersESC6: Insufficient powerESC7: Technical failuresESC8: Terrorist attacks
Energy supply source preferenceESP1: CoalESP2: Natural gasESP3: Hydroelectric powerESP4: Nuclear powerESP5: Sun or solar powerESP6: Wind powerESP7: Biomass energy
Personal normsPN: Personal responsibility to reduce climate change
Summary
Psychological networks are useful for:
• Exploratory analyses of (relatively) high-dimensional data• Validation of scales of various items• Testing hypotheses on the absence/presence, direction or strength ofrelations
• Testing hypotheses on the co-occurrence of relations
With these models, we explored and confirmed theories on energy-relatedbehaviour and found differences and similarities between countries.
References & credits
Papers in preparation on this topic:
• N. Bhushan, F. Mohnert, C. J. Albers, L. Jans, L. Steg. The value of Gaussian graphicalmodels to explore relationships between environmental psychological constructs.
• M. Verschoor, C. J. Albers, L. Steg. A network model of the environmental module inthe European Social Survey 2016
• T. Bouman, M. Verschoor, L. Steg, G. Böhm, S. D. Fisher, W. Poortinga, L. Whitmarsh, C.J. Albers. General personal factors predicting energy-saving behaviours and climatepolicy support
• E. J. Sharpe, M. Verschoor, L. Steg, G. Perlaviciute, C. J. Albers. Similarity encouragesconsistency across energy-saving behaviour and energy policy support in Europeand beyond
Further reading – Environmental Psychology
• Abrahamse, W. (2007). Energy conservation through behavioural change. Universityof Groningen
• Van der Werff, E., Steg, L., & Keizer, K. (2013). The value of environmental self-identity:The relationship between biospheric values, environmental self-identity andenvironmental preferences, intentions and behaviour. Journal of EnvironmentalPsychology.
• Steg, L., Bolderdijk, J.W., Keizer, K., & Perlaviciute, G.(2014). An integrated frameworkfor encouraging pro-environmental behaviour: the role of values, situational factorsand goals. Journal of Environmental Psychology ofvalues,situationalfactorsan
• Bamberg, S., Rees, J., & Seebauer, S. (2015). Collective climate action: Determinants ofparticipation intention in community-based pro-environmental initiatives. Journal ofEnvironmental Psychology.
• Steg, L., Perlaviciute, G., & van der Werff, E. (2015). Understanding the humandimensions of a sustainable energy transition. Frontiers in Psychology.
Further reading – Psychological Networks
• Dempster, A.P. (1972). Covariance selection. Biometrics.
• Lauritzen, S.L. (1996). Graphical Models. Oxford University Press.
• Snijders, T.A.B. (2002). The statistical evaluation of social network dynamics.Sociological methodology.
• Opsahl, T., Skvoretz, J. (2010). Node centrality in weighted networks. Social networks.
• Cramer, A.O.J., Waldorp, L., van der Maas, H., & Borsboom, D. (2010). Comorbidity: ANetwork Perspective. Behavioral and Brain Sciences.
• Borsboom, D., & Cramer, A. O. (2013). Network analysis: an integrative approach tothe structure of psychopathology. Annual review of clinical psychology.
• Bringmann, L.F. (2016). Dynamical networks in psychology: More than a prettypicture? University of Leuven.
• Epskamp, S., Borsboom, D., & Fried, E. I. (2017). Estimating psychological networksand their accuracy: a tutorial paper. Behavioural Research.